I guess if determinism is so important implementations can be changed, it is just maybe not that high priority.
8,476 karma · joined September 7, 2018
I guess if determinism is so important implementations can be changed, it is just maybe not that high priority.
At the same time if you can think about ending yourself for something light and fun, you should go to psychologist to process why it's the case.
Every 2 hours a dirty cruise ship comes there, hundreds of people rushing through the stairs to take their photos then going back on the same (narrow) passage where everybody usually travels.
They don't have time for trying out the amazing restaurants, walks and hotels it provides.
It's a way of saying ,,I give you a loan but I don't tell you how much more money you have to pay me back, maybe 5 times more interest as you may think now''.
If I watch only 3 Netflix movies or 1 serie in a month, I should be getting 10 times more budget per movie / serie than a person viewing 30 Netflix movies / 10 series. Sadly this doesn't happen.
The way to implement anything new and cool is to read it all first.
I love the way convolutions and matmuls are both expressed using ShapeTracker and simple mul operation for example. No other framework has that elegance.
Until you understand that part, it's not worth trying to change it.
Try adding a new feature to Tensorflow / PyTorch, do the same for TinyGrad and you won't have this question anymore.
We also know the solutions: electric cars (I just bought my first), heat pumps with electric heating, and reversing aging.
I don't want to see interviews about you quitting and saving the world, I wanted to see how you want to make money, and finally this last was honest. I guess you had to be quiet before, but I guess we had enough ,,altruisms'' and congress hearings with a few Sams already.
(I also have Aspergers though, so not representing most people)
If you look at the series, they took 14 random people for each group and gave them 2 days of survival training and basic helper tools.
Just looking at how the men killed the crocodile in a team work with their bare hands first time in their life without any help from experts looked awesome.
The most interesting part was not how much women failed to survive, and how successful men were (which makes sense, as men are designed for survival), but the feeling of how much easier in that world is to show the need for masculanity (which is mostly not needed anymore).
Generally in a group of healthy young population with 50% sex ratio, about half of the men are selected, and they help eachother.
But in the online dating / birth control age this ratio has gone more extreme, and there's no real ,,parenting advice'' or emotional need support that can help, just accepting the new normal for everyone, and competing harder.
My tablet and phone is ARM as well, and I'm not considering switching them to intel either.
I don't know about U.S., but in my country there's an administrative team that has to approve gender change (operations are not required by law).
It's easy to debug because the generated kernels can be compared to GGML, and still gives something practical that we all can play with.
At this point breadth first is a bit boring, because this way we don't know how far tinygrad is from optimal generated output.
That's why we're stuck with lifetimes as function contracts.
So if you want to try something like that, just look at Go :)
In the case of automatically upgrading from int to bignum, or green threads, going lower level would be much harder.
I just wish Rust stayed focused on being the best systems level programming language (for example finishing the SIMD package, getting into stable Rust, getting language level GPU integration similar to CUDA / OpenCL).
PyToch, and now the newer GGML for example is still written in C++ for example
I still like the changes being made in the language, just not the focus.